Add Python-style list-comprehension syntax support to SQL

[Mytherin]

The following syntax is now supported:

SELECT [x + 1 for x in [1, 2, 3]] AS l;
┌───────────┐
│     l     │
├───────────┤
│ [2, 3, 4] │
└───────────┘
SELECT [x + 1 for x in [1, 2, 3] if x >= 2] AS l;
┌────────┐
│   l    │
├────────┤
│ [3, 4] │
└────────┘

Under the hood this just translates to the list_apply and list_filter functions as implemented in #3694. Any functions or expressions can be used for both the filter and the main expression. For example:

CREATE TABLE store_inventory(store_id INTEGER, fruits VARCHAR[]);
INSERT INTO store_inventory VALUES (1, ['Apple', 'Banana', 'Mango']);
INSERT INTO store_inventory VALUES (2, ['Mango', 'Pineapple']);
SELECT [concat(store_id, '-', fruit) for fruit in fruits if lower(fruit) IN ('mango', 'banana')] AS filtered_fruit
FROM store_inventory;
┌─────────────────────┐
│   filtered_fruit    │
├─────────────────────┤
│ [1-Banana, 1-Mango] │
│ [2-Mango]           │
└─────────────────────┘

[Alex-Monahan]

This is awesome! We might need an "Even friendlier SQL with DuckDB" blog post at this rate!!

[hannes]

We can also just add an Update section to the blog post for now